hugoycj/Instant-angelo
Neuralangelo-based 3D surface reconstruction system that creates photorealistic digital twins from mobile device video captures.

Instant-angelo accelerates high-fidelity neural 3D surface reconstruction using an enhanced Neuralangelo implementation. It takes RGB video captured from mobile devices and produces detailed digital twins with accurate geometry and appearance. The project includes a reimplementation of UniSDF for improved reflective surface reconstruction and optimizes the pipeline to achieve results in approximately 20 minutes on standard GPU resources.
Frequently asked
- What is hugoycj/Instant-angelo?
- Neuralangelo-based 3D surface reconstruction system that creates photorealistic digital twins from mobile device video captures.
- Is Instant-angelo open source?
- Yes — hugoycj/Instant-angelo is open source, released under the MIT license.
- What language is Instant-angelo written in?
- hugoycj/Instant-angelo is primarily written in Python.
- How popular is Instant-angelo?
- hugoycj/Instant-angelo has 462 stars on GitHub.
- Where can I find Instant-angelo?
- hugoycj/Instant-angelo is on GitHub at https://github.com/hugoycj/Instant-angelo.